히브리어에서 페르시아어으로 번역
HE FA
즉각적인 결과
사용 가능한 API
Hebrew에서 Persian로 번역하는 방법
1
텍스트 입력
위의 입력 상자에 Hebrew 텍스트를 입력하거나 붙여넣으십시오. 최대 10,000자를 입력하거나 문서 파일을 업로드할 수 있습니다.
2
번역을 클릭합니다
번역 버튼을 클릭하거나 단순히 기다리십시오. 입력을 중지하면 자동 번역이 시작됩니다. AI가 밀리초 안에 텍스트를 처리합니다.
3
귀하의 번역을 복사
Persian 번역이 즉시 나타납니다. 복사 버튼을 클릭하여 클립보드에 복사하거나 API를 통해 통합하십시오.
일반적인 Hebrew 구문
즉시 번역 어떤 구문을 클릭하십시오.
인스턴트 번역
밀리초 내에 Hebrew에서 Persian으로 번역을 가져옵니다.
문서 지원
일괄 번역을 위해 Word, PDF, SRT 자막 등을 업로드하십시오.
API 액세스
REST API를 사용하여 이 번역 쌍을 앱에 통합하십시오.
curl -X POST https://api.translateapi.ai/api/v1/translate/ \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"text": "Hello", "source_language": "he", "target_language": "fa"}'Hebrew에서 Persian으로 번역하기
TranslateAPI는 고급 AI로 구동되는 빠르고 정확한 Hebrew에서 Persian로의 번역을 제공합니다.
- 텍스트 번역 - 최대 10,000자의 텍스트를 즉시 번역
- 문서 번역 - 번역용 Word, PDF, 텍스트 파일 업로드
- API 통합 - 앱에 Hebrew에서 Persian으로 번역 추가
- 일괄 번역 - 단일 요청에서 여러 텍스트 번역
사용 사례:
- 비즈니스를 위해 Hebrew 문서를 Persian로 번역
- Hebrew에서 Persian으로 웹사이트 및 앱 현지화
- Hebrew 자막을 Persian 로 변환
- Persian 스피커와 의사소통하기
번역 쌍
- 소스
- 히브리어 (
he) - 대상
- 페르시아어 (
fa)
자주 묻는 질문
When both source and target are right-to-left, the model translates without inserting any left-to-right segments unless the source had Latin content (brand names, URLs) — those islands keep their original direction and are bracketed with the appropriate bidi control codes.
No — Arabic, Hebrew, Persian, and Urdu use related script families but the languages themselves are unrelated. Hebrew to Persian is a full neural translation, not transliteration. The model is trained on bilingual corpora that pair the two languages directly where available, with English as a fallback bridge for rarer pair combinations.
Loanwords are translated to their native equivalent in Persian (e.g. "kitab" → the Hebrew "sefer" when going Arabic → Hebrew) unless the source clearly used the loanword as a name or quotation. Brand names always stay phonetic.
Hebrew is unicameral and uses block letterforms; Arabic and Urdu are cursive with contextual shaping (initial/medial/final/isolated). The output uses the correct script with its native rendering rules — your editor or browser handles the shaping automatically as long as the font supports it.
Vowel marks tend to be lost in translation between RTL languages because each script encodes them differently and the target writing convention often omits them. If you need fully-vocalized output for Persian, post-process with a vocalizer specific to that language.
Both RTL languages default to the literary / formal register (MSA for Arabic, Modern Israeli Hebrew literary, Persian rasmi, Urdu adabi). Conversational output between two RTL targets often needs a native review pass because each language has independent register conventions.
Output between two RTL languages uses the standard literary form of the target — dialect-to-dialect translation (e.g. Egyptian Arabic colloquial → spoken Persian Tehrani) is not directly supported. Translate to MSA / standard target first, then localize.
Quoted scripture (Qur'anic verses, Hebrew Bible, hadith) is preserved verbatim with original orthography when wrapped in quotation marks. Surrounding commentary is translated normally.
Calendar terms translate to the Persian convention (Gregorian month names if Persian commonly uses them, Hijri / Hebrew calendar terms where contextually appropriate). Numerals default to Latin digits in the output.
Up to 10,000 characters via the web translator and 50,000 via the API. Long-form documents in RTL pairs benefit from paragraph-sized chunks (300-1000 characters) because the model maintains discourse coherence within a chunk.
Yes — Latin segments inside an RTL document keep their left-to-right direction and Latin spelling. The model only translates the RTL prose; embedded URLs, emails, code blocks, and Latin brand names pass through unchanged.
Yes — upload an image or PDF of Hebrew text and the OCR pipeline applies script-aware preprocessing so right-to-left column-and-line order is detected before translation to Persian. Vowel marks usually do not survive OCR; expect un-vocalized output.